Field of Invention
[0001] The present invention relates to data processing apparatus operable to map input
symbols onto sub-carrier signals of an Orthogonal Frequency Division Multiplexed (OFDM)
symbol. The present invention also relates to an address generator for use in writing
symbols to/reading symbols from an interleaver memory.
[0002] Embodiments of the present invention can provide an OFDM transmitter.
Background of the Invention
[0003] The Digital Video Broadcasting-Terrestrial standard (DVB-T) utilises Orthogonal Frequency
Division Multiplexing (OFDM) to communicate data representing video images and sound
to receivers via a broadcast radio communications signal. There are known to be two
modes for the DVB-T standard which are known as the 2k and the 8k mode. The 2k mode
provides 2048 sub-carriers whereas the 8k mode provides 8192 sub-carriers. Similarly
for the Digital Video Broadcasting-Handheld standard (DVB-H) a 4k mode has been provided,
in which the number of sub-carriers is 4096.
[0004] In order to improve the integrity of data communicated using DVB-T or DVB-H a symbol
interleaver is provided in order to interleave input data symbols as these symbols
are mapped onto the sub-carrier signals of an OFDM symbol. Such a symbol interleaver
comprises an interleaver memory in combination with an address generator. The address
generator generates an address for each of the input symbols, each address indicating
one of the sub-carrier signals of the OFDM symbol onto which the data symbol is to
be mapped. For the 2k mode and the 8k mode an arrangement has been disclosed in the
DVB-T standard for generating the addresses for the mapping. Likewise for the 4k mode
of DVB-H standard, an arrangement for generating addresses for the mapping has been
provided and an address generator for implementing this mapping is disclosed in European
Patent application
04251667.4. The address generator comprises a linear feed back shift register which is operable
to generate a pseudo random bit sequence and a permutation circuit. The permutation
circuit permutes the order of the content of the linear feed back shift register in
order to generate an address. The address provides an indication of one of the OFDM
sub-carriers for carrying an input data symbol stored in the interleaver memory, in
order to map the input symbols onto the sub-carrier signals of the OFDM symbol.
[0005] In accordance with a further development of the Digital Video Broadcasting-Terrestrial
broadcasting standard, known as DVB-T2 there has been proposed that further modes
for communicating data be provided.
Summary of Invention
[0006] According to an aspect of the present invention there is provided a data processing
apparatus operable to map input data symbols to be communicated onto a predetermined
number of sub-carrier signals of an Orthogonal Frequency Division Multiplexed (OFDM)
symbol. The data processing apparatus comprises an interleaver operable to read-into
a memory the predetermined number of data symbols for mapping onto the OFDM sub-carrier
signals, and to read-out of the memory the data symbols for the OFDM sub-carriers
to effect the mapping. The read-out is in a different order than the read-in, the
order being determined from a set of addresses, with the effect that the data symbols
are interleaved on the sub-carrier signals. The set of addresses is determined by
an address generator, an address being generated for each of the input symbols to
indicate one of the sub-carrier signals onto which the data symbol is to be mapped.
[0007] The address generator comprises a linear feedback shift register including a predetermined
number of register stages and is operable to generate a pseudo-random bit sequence
in accordance with a generator polynomial, and a permutation circuit and a control
unit. The permutation circuit is operable to receive the content of the shift register
stages and to permute the bits present in the register stages in accordance with a
permutation order to form an address of one of the OFDM sub-carriers.
[0008] The control unit is operable in combination with an address check circuit to re-generate
an address when a generated address exceeds a predetermined maximum valid address.
The data processing apparatus is characterised in that the predetermined maximum valid
address is approximately thirty two thousand, the linear feedback shift register has
fourteen register stages with a generator polynomial for the linear feedback shift
register of

and the permutation order forms, with an additional bit, a fifteen bit address
Ri[
n] for the i-th data symbol from the bit present in the n-th register stage

in accordance with the table:
| R'i bit positions |
13 |
12 |
11 |
10 |
9 |
8 |
7 |
6 |
5 |
4 |
3 |
2 |
1 |
0 |
| Ri bit positions |
6 |
5 |
0 |
10 |
8 |
1 |
11 |
12 |
2 |
9 |
4 |
3 |
13 |
7 |
[0009] Although it is known within the DVB-T standard to provide the 2k mode and the 8k
mode, and the DVB-H standard provides a 4k mode, there has been proposed to provide
a 32k mode for DVB-T2. Whilst the 8k mode provides an arrangement for establishing
a single frequency network with sufficient guard periods to accommodate larger propagation
delays between DVB transmitters, the 2k mode is known to provide an advantage in mobile
applications. This is because the 2k symbol period is only one quarter of the 8k symbol
period, allowing the channel estimation to be more frequently updated allowing the
receiver to track the time variation of the channel due to doppler and other effects
more accurately. The 2k mode is therefore advantageous for mobile applications.
[0010] In order to provide an even sparser deployment of DVB transmitters within a single
frequency network, it has been proposed to provide the 32k mode. To implement the
32k mode, a symbol interleaver must be provided for mapping the input data symbols
onto the sub-carrier signals of the OFDM symbol.
[0011] Embodiments of the present invention can provide a data processing apparatus operable
as a symbol interleaver for mapping data symbols to be communicated on an OFDM symbol,
having approximately thirty two thousand sub-carrier signals. In one embodiment the
number of sub-carrier signals maybe a value substantially between twenty four thousand
and thirty two thousand seven hundred and sixty eight. Furthermore, the OFDM symbol
may include pilot sub-carriers, which are arranged to carry known symbols, and the
predetermined maximum valid address depends on a number of the pilot sub-carrier symbols
present in the OFDM symbol. As such the 32k mode can be provided for example for a
DVB standard, such as DVB-T2, DVB-T or DVB-H.
[0012] Mapping data symbols to be transmitted onto the sub-carrier signals of an OFDM symbol,
where the number of sub-carrier signals is approximately thirty two thousand, represents
a technical problem requiring simulation analysis and testing to establish an appropriate
generator polynomial for the linear feedback shift register and the permutation order.
This is because the mapping requires that the symbols are interleaved onto the sub-carrier
signals with the effect that successive symbols from the input data stream are separated
in frequency by a greatest possible amount in order to optimise the performance of
error correction coding schemes.
[0013] Error correction coding schemes such as LDPC/BCH coding, which has been proposed
for DVB-T2 perform better when noise and degradation of the symbol values resulting
from communication is un-correlated. Terrestrial broadcast channels may suffer from
correlated fading in both the time and the frequency domains. As such by separating
encoded symbols on to different sub-carrier signals of the OFDM symbol by as much
as possible, the performance of error correction coding schemes can be increased.
[0014] As will be explained, it has been discovered from simulation performance analysis
that the generator polynomial for the linear feed back shift register in combination
with the permutation circuit order indicated above, provides a good performance. Furthermore,
by providing an arrangement which can implement address generating for each of the
2k mode, the 4k mode and the 8k mode by changing the taps of the generator polynomial
for the linear feed back shift register and the permutation order, a cost effective
implementation of the symbol interleaver for the 32k mode can be provided. Furthermore,
a transmitter and a receiver can be changed between the 2k mode, 4k mode, 8k mode
and the 32k mode by changing the generator polynomial and the permutation orders.
This can be effected in software (or by embedded signalling) whereby a flexible implementation
is provided. Alternatively, the 32k mode interleaver can be used with the 2k, 4k,
8k and a 16k mode by simply adapting the maximum valid address in the address generator
in accordance with the number of data bearing sub-carriers per OFDM symbol.
[0015] The additional bit, which is used to form the address from the content of the linear
feedback shift register, may be produced by a toggle circuit, which changes from 1
to 0 for each address, so as to reduce a likelihood that if an address exceeds the
predetermined maximum valid address, then the next address will be a valid address.
In one example the additional bit is the most significant bit.
[0016] Various aspects and features of the present invention are defined in the appended
claims. Further aspects of the present invention include a data processing apparatus
operable to map symbols received from a predetermined number of sub-carrier signals
of an Orthogonal Frequency Division Multiplexed (OFDM) symbol into an output symbol
stream, as well as a transmitter and a receiver.
Brief Description of Drawings
[0017] Embodiments of the present invention will now be described by way of example only
with reference to the accompanying drawings, wherein like parts are provided with
corresponding reference numerals, and in which:
Figure 1 is a schematic block diagram of a Coded OFDM transmitter which may be used,
for example, with the DVB-T2 standard;
Figure 2 is a schematic block diagram of parts of the transmitter shown in Figure
1 in which a symbol mapper and a frame builder illustrate the operation of an interleaver;
Figure 3 is a schematic block diagram of the symbol interleaver shown in Figure 2;
Figure 4 is a schematic block diagram of an interleaver memory shown in Figure 3 and
the corresponding symbol de-interleaver in the receiver;
Figure 5 is a schematic block diagram of an address generator shown in Figure 3 for
the 32k mode;
Figure 6(a) is diagram illustrating results for an interleaver using the address generator
shown in Figure 5 for even symbols and Figure 6(b) is a diagram illustrating design
simulation results for odd symbols, whereas Figure 6(c) is a diagram illustrating
comparative results for an address generator using a different permutation code for
even and Figure 6(d) is a corresponding diagram for odd symbols;
Figure 7 is a schematic block diagram of a Coded OFDM receiver which may be used,
for example, with the DVB-T2 standard; and
Figure 8 is a schematic block diagram of a symbol de-interleaver which appears in
Figure 7.
Description of Preferred Embodiments
[0018] It has been proposed that the number of modes, which are available within the DVB-T2
standard should be extended to include a 1k mode, a 16k mode and a 32k mode. The following
description is provided to illustrate the operation of a symbol interleaver in accordance
with the present technique, although it will be appreciated that the symbol interleaver
can be used with other modes and other DVB standards.
[0019] Figure 1 provides an example block diagram of a Coded OFDM transmitter which may
be used for example to transmit video images and audio signals in accordance with
the DVB-T2 standard. In Figure 1 a program source generates data to be transmitted
by the COFDM transmitter. A video coder 2, and audio coder 4 and a data coder 6 generate
video, audio and other data to be transmitted which are fed to a program multiplexer
10. The output of the program multiplexer 10 forms a multiplexed stream with other
information required to communicate the video, audio and other data. The multiplexer
10 provides a stream on a connecting channel 12. There may be many such multiplexed
streams which are fed into different branches A, B etc. For simplicity, only branch
A will be described.
[0020] As shown in Figure 1 a COFDM transmitter 20 receives the stream at a multiplexer
adaptation and energy dispersal block 22. The multiplexer adaptation and energy dispersal
block 22 randomises the data and feeds the appropriate data to a forward error correction
encoder 24 which performs error correction encoding of the stream. A bit interleaver
26 is provided to interleave the encoded data bits which for the example of DVB-T2
is the LDCP/BCH encoder output. The output from the bit interleaver 26 is fed to a
bit into constellation mapper 28, which maps groups of bits onto a constellation point,
which is to be used for conveying the encoded data bits. The outputs from the bit
into constellation mapper 28 are constellation point labels that represent real and
imaginary components. The constellation point labels represent data symbols formed
from two or more bits depending on the modulation scheme used. These will be referred
to as data cells. These data cells are passed through a time-interleaver 30 whose
effect is to interleaver data cells resulting from multiple LDPC code words.
[0021] The data cells are received by a frame builder 32, with data cells produced by branch
B etc in Figure 1, via other channels 31. The frame builder 32 then forms many data
cells into sequences to be conveyed on COFDM symbols, where a COFDM symbol comprises
a number of data cells, each data cell being mapped onto one of the sub-carriers.
The number of sub-carriers will depend on the mode of operation of the system, which
may include one of 1k, 2k, 4k, 8k, 16k or 32k, each of which provides a different
number of sub-carriers according, for example to the following table:
Number of Sub-carriers Adapted from DVB-T/H
| Mode |
Sub-carriers |
| 1K |
756 |
| 2K |
1512 |
| 4K |
3024 |
| 8K |
6048 |
| 16K |
12096 |
| 32K |
24192 |
[0022] Thus in one example, the number of sub-carriers for the 32k mode is twenty four thousand
one hundred and ninety two. For the DVB-T2 system, the number of sub-carriers per
OFDM symbol can vary depending upon the number of pilot and other reserved carriers.
Thus, in DVB-T2, unlike in DVB-T, the number of sub-carriers for carrying data is
not fixed. Broadcasters can select one of the operating modes from 1k, 2k, 4k, 8k,
16k, 32k each providing a range of sub-carriers for data per OFDM symbol, the maximum
available for each of these modes being 1024, 2048, 4096, 8192, 16384, 32768 respectively.
In DVB-T2 a physical layer frame is composed of many OFDM symbols. Typically the frame
starts with one or more preamble or P2 OFDM symbols, which are then followed by a
number payload carrying OFDM symbols. The end of the physical layer frame is marked
by a frame closing symbols. For each operating mode, the number of sub-carriers may
be different for each type of symbol. Furthermore, this may vary for each according
to whether bandwidth extension is selected, whether tone reservation is enabled and
according to which pilot sub-carrier pattern has been selected. As such a generalisation
to a specific number of sub-carriers per OFDM symbol is difficult. However, the frequency
interleaver for each mode can interleave any symbol whose number of sub-carriers is
smaller than or the same as the maximum available number of sub-carriers for the given
mode. For example, in the 1k mode, the interleaver would work for symbols with the
number of sub-carriers being less than or equal to 1024 and for 16k mode, with the
number of sub-carriers being less than or equal to 16384.
[0023] The sequence of data cells to be carried in each COFDM symbol is then passed to the
symbol interleaver 33. The COFDM symbol is then generated by a COFDM symbol builder
block 37 which introduces pilot and synchronising signals fed from a pilot and embedded
signal former 36. An OFDM modulator 38 then forms the OFDM symbol in the time domain
which is fed to a guard insertion processor 40 for generating a guard interval between
symbols, and then to a digital to analogue convertor 42 and finally to an RF amplifier
within an RF frontend 44 for eventual broadcast by the COFDM transmitter from an antenna
46.
Providing a 32k Mode
[0024] To create a new 32K mode, several elements are to be defined, one of which is the
32K symbol interleaver33. The bit to constellation mapper 28, symbol interleaver 33
and the frame builder 32 are shown in more detail in Figure 2.
[0025] As explained above, the present invention provides a facility for providing a quasi-optimal
mapping of the data symbols onto the OFDM sub-carrier signals. According to the example
technique the symbol interleaver is provided to effect the optimal mapping of input
data symbols onto COFDM sub-carrier signals in accordance with a permutation code
and generator polynomial, which has been verified by simulation analysis.
[0026] As shown in Figure 2 a more detailed example illustration of the bit to symbol constellation
mapper 28 and the frame builder 32 is provided to illustrate an example embodiment
of the present technique. Data bits received from the bit interleaver 26 via a channel
62 are grouped into sets of bits to be mapped onto a data cell, in accordance with
a number of bits per symbol provided by the modulation scheme. The groups of bits,
which forms a data word, are fed in parallel via data channels 64 the a mapping processor
66. The mapping processor 66 then selects one of the data symbols, in accordance with
a pre-assigned mapping. The constellation point, is represented by a real and an imaginary
component is provided to the output channel 29 as one of a set of inputs to the frame
builder 32.
[0027] The frame builder 32 receives the data cells from the bit to constellation mapper
28 through channel 29, together with data cells from the other channels 31. After
building a frame of many COFDM cell sequences, the cells of each COFDM symbol are
then written into an interleaver memory 100 and read out of the interleaver memory
100 in accordance with write addresses and read addresses generated by an address
generator 102. According to the write-in and read-out order, interleaving of the data
cells is achieved, by generating appropriate addresses. The operation of the address
generator 102 and the interleaver memory 100 will be described in more detail shortly
with reference to Figures 3, 4 and 5. The interleaved data cells are then combined
with pilot and synchronisation symbols received from the pilot and embedded signalling
former 36 into an OFDM symbol builder 37, to form the COFDM symbol, which is fed to
the OFDM modulator 38 as explained above.
Interleaver
[0028] Figure 3 provides an example of parts of the symbol interleaver 33, which illustrates
the present technique for interleaving symbols. In Figure 3 the input data cells from
the frame builder 32 are written into the interleaver memory 100. The data cells are
written into the interleaver memory 100 according to a write address fed from the
address generator 102 on channel 104, and read out from the interleaver memory 100
according to a read address fed from the address generator 102 on a channel 106. The
address generator 102 generates the write address and the read address as explained
below, depending on whether the COFDM symbol is odd or even, which is identified from
a signal fed from a channel 108, and depending on a selected mode, which is identified
from a signal fed from a channel 110. As explained, the mode can be one of a 1k mode,
2k mode, 4k mode, 8k mode, 16k mode or a 32k mode. As explained below, the write address
and the read address are generated differently for odd and even symbols as explained
with reference to Figure 4, which provides an example implementation of the interleaver
memory 100.
[0029] In the example shown in Figure 4, the interleaver memory is shown to comprise an
upper part 100 illustrating the operation of the interleaver memory in the transmitter
and a lower part 340, which illustrates the operation of the de-interleaver memory
in the receiver. The interleaver 100 and the de-interleaver 340 are shown together
in Figure 4 in order to facilitate understanding of their operation. As shown in Figure
4 a representation of the communication between the interleaver 100 and the de-interleaver
340 via other devices and via a transmission channel has been simplified and represented
as a section 140 between the interleaver 100 and the de-interleaver 340. The operation
of the interleaver 100 is described in the following paragraphs:
[0030] Although Figure 4 provides an illustration of only four input data cells onto an
example of four sub-carrier signals of a COFDM symbol, it will be appreciated that
the technique illustrated in Figure 4 can be extended to a larger number of sub-carriers
such as 756 for the 1k mode 1512 for the 2k mode, 3024 for the 4k mode and 6048 for
the 8k mode, 12096 for the 16k mode and 24192 for the 32k mode.
[0031] The input and output addressing of the interleaver memory 100 shown in Figure 4 is
shown for odd and even symbols. For an even COFDM symbol the data cells are taken
from the input channel 77 and written into the interleaver memory 124.1 in accordance
with a sequence of addresses 120 generated for each COFDM symbol by the address generator
102. The write addresses are applied for the even symbol so that as illustrated interleaving
is effected by the shuffling of the write-in addresses. Therefore, for each interleaved
symbol y(h(q)) = y'(q).
[0032] For odd symbols the same interleaver memory 124.2 is used. However, as shown in Figure
4 for the odd symbol the write-in order 132 is in the same address sequence used to
read out the previous even symbol 126. This feature allows the odd and even symbol
interleaver implementations to only use one interleaver memory 100 provided the read-out
operation for a given address is performed before the write-in operation. The data
cells written into the interleaver memory 124 during odd symbols are then read out
in a sequence 134 generated by the address generator 102 for the next even COFDM symbol
and so on. Thus only one address is generated per symbol, with the read-in and write-out
for the odd/even COFDM symbol being performed contemporaneously.
[0033] In summary, as represented in Figure 4, once the set of addresses H(q) has been calculated
for all active sub-carriers, the input vector Y' = (y
0', y
1', y
2',...y
Nmax-1') is processed to produce the interleaved vector Y = (y
0, y
1, y
2,...y
Nmax-1) defined by:
y
H(q) = y'
q for even symbols for q = 0,...,N
max-1
y
q = y'
H(q) for odd symbols for q = 0,...,N
max-1
[0034] In other words, for even OFDM symbols the input words are written in a permutated
way into a memory and read back in a sequential way, whereas for odd symbols, they
are written sequentially and read back permutated. In the above case, the permutation
H(q) is defined by the following table:
Table 1: permutation for simple case where Nmax =4
[0035] As shown in Figure 4, the de-interleaver 340 operates to reverse the interleaving
applied by the interleaver 100, by applying the same set of addresses as generated
by an equivalent address generator, but applying the write-in and read-out addresses
in reverse. As such, for even symbols, the write-in addresses 342 are in sequential
order, whereas the read out address 344 are provided by the address generator. Correspondingly,
for the odd symbols, the write-in order 346 is determined from the set of addresses
generated by the address generator, whereas read out 348 is in sequential order.
Address Generation for the 32k Mode
[0036] A schematic block diagram of the algorithm used to generate the permutation function
H(q) is represented in Figure 5 for the 32K mode.
[0037] An implementation of the address generator 102 for the 32k mode is shown in Figure
5. In Figure 5 a linear feed back shift register is formed by thirteen register stages
200 and a xor-gate 202 which is connected to the stages of the shift register 200
in accordance with a generator polynomial. Therefore, in accordance with the content
of the shift register 200 a next bit of the shift register is provided from the output
of the xor-gate 202 by xoring the content of shift registers R[0], R[1], R[2], R[12]
according to the generator polynomial:

[0038] According to the generator polynomial a pseudo random bit sequence is generated from
the content of the shift register 200. However, in order to generate an address for
the 32k mode as illustrated, a permutation circuit 210 is provided which effectively
permutes the order of the bits within the shift register 200.1 from an order
R'i[n] to an order
R¡[n] at the output of the permutation circuit 210. Fourteen bits from the output of the
permutation circuit 210 are then fed on a connecting channel 212 to which is added
a most significant bit via a channel 214 which is provided by a toggle circuit 218.
A fifteen bit address is therefore generated on channel 212. However, in order to
ensure the authenticity of an address, an address check circuit 216 analyses the generated
address to determine whether it exceeds a predetermined maximum value. The predetermined
maximum value may correspond to the maximum number of sub-carrier signals, which are
available for data symbols within the COFDM symbol, available for the mode which is
being used. However, the interleaver for the 32k mode may also be used for other modes,
so that the address generator 102 may also be used for the 2k mode, 4k mode, 8k mode,
16k mode and the 32k mode, by adjusting accordingly the number of the maximum valid
address.
[0039] If the generated address exceeds the predetermined maximum value then a control signal
is generated by the address check unit 216 and fed via a connecting channel 220 to
a control unit 224. If the generated address exceeds the predetermined maximum value
then this address is rejected and a new address regenerated for the particular symbol.
[0040] For the 32k mode, an (N
r - 1) bit word R'
i is defined, with N
r = log
2 M
max, where M
max = 32768 using a LFSR (Linear Feedback Shift Register).
[0041] The polynomials used to generate this sequence is:

where i varies from 0 to M
max - 1
[0042] Once one R'
i, word has been generated, the R'
i, word goes through a permutation to produce another (N
r - 1) bit word called R
i. R
i is derived from R'
i by the bit permutations given as follows:
| R'i bit positions |
13 |
12 |
11 |
10 |
9 |
8 |
7 |
6 |
5 |
4 |
3 |
2 |
1 |
0 |
| Ri bit positions |
6 |
5 |
0 |
10 |
8 |
1 |
11 |
12 |
2 |
9 |
4 |
3 |
13 |
7 |
[0043] Bit permutation for the 32K mode
[0044] As an example, this means that for the mode 32K, the bit number 12 of R'
i is sent to bit position number 5 of R
i.
[0045] The address H(q) is then derived from R
i through the following equation:

[0046] The (i mod2) · 2
Nr-1 part of the above equation is represented in Figure 5 by the toggle block T 218.
[0047] An address check is then performed on H(q) to verify that the generated address is
within the range of acceptable addresses: if (H(q)<N
max), where N
max =24192, for example, in the 32K mode, then the address is valid. If the address is
not valid, the control unit is informed and it will try to generate a new H(q) by
incrementing the index i.
[0048] The role of the toggle block is to make sure that we do not generate an address exceeding
N
max twice in a row. In effect, if an exceeding value was generated, this means that the
MSB (i.e. the toggle bit) of the address H(q) was one. So the next value generated
will have a MSB set to zero, insuring to produce a valid address.
[0049] The following equations sum up the overall behaviour and help to understand the loop
structure of this algorithm:

Analysis Supporting the Address Generator for the 32k Mode
[0050] The selection of the polynomial generator and the permutation code explained above
for the address generator 102 for the 1k mode has been identified following simulation
analysis of the relative performance of the interleaver. The relative performance
of the interleaver has been evaluated using a relative ability of the interleaver
to separate successive symbols or an "interleaving quality". As mentioned above, effectively
the interleaving must perform for both odd and even symbols, in order to use a single
interleaver memory. The relative measure of the interleaver quality is determined
by defining a distance D (in number of sub-carriers).. A criterion C is chosen to
identify a number of sub-carriers that are at distance ≤ D at the output of the interleaver
that were at distance ≤ D at the input of the interleaver, the number of sub-carriers
for each distance D then being weighted with respect to the relative distance. The
criterion C is evaluated for both odd and even COFDM symbols. Minimising C produces
a superior quality interleaver.

where: N
even(d) and N
odd(d) are number of sub-carriers in an even and odd symbol respectively at the output
of the interleaver that remain within d sub-carrier spacing of each other.
[0051] Analysis of the interleaver identified above for the 32k mode for a value of D =
5 is shown in Figure 6(a) for the even COFDM symbols and in Figure 6(b) for the odd
COFDM symbol. According to the above analysis, the value of C for the permutation
code identified above for the 32k mode produced a value of C = 21.75, that the weighted
number of sub-carriers with symbols which are separated by five or less in the output
according to the above equation was 21.75.
[0052] A corresponding analysis is provided for an alternative permutation code for even
COFDM symbols in Figure 6(c) for odd COFDM symbols in Figure 6(d). As can be seen
in comparison to the results illustrated in Figures 6(a) and 6(b), there are more
components present which represent symbols separated by small distances such as D
= 1, and D = 2, when compared with the results shown in Figure 6(a) and 6(b), illustrating
that the permutation code identified above for the 32k mode symbol interleaver produces
a superior quality interleaver.
Alternative Permutation Codes
[0053] The following fifteen alternative possible codes ([n]R
i bit positions, where n = 1 to 15) have been found to provide a symbol interleaver
with a good quality as determined by the criterion C identified above.
| R'i bit positions |
13 |
12 |
11 |
10 |
9 |
8 |
7 |
6 |
5 |
4 |
3 |
2 |
1 |
0 |
| [1]Ri bit positions |
0 |
6 |
1 |
7 |
2 |
11 |
12 |
5 |
9 |
8 |
3 |
10 |
4 |
13 |
| [2]Ri bit positions |
9 |
5 |
0 |
7 |
2 |
8 |
3 |
6 |
12 |
11 |
4 |
1 |
10 |
13 |
| [3]Ri bit positions |
9 |
12 |
0 |
1 |
2 |
13 |
5 |
8 |
6 |
3 |
7 |
4 |
10 |
11 |
| [4]Ri bit positions |
13 |
8 |
1 |
12 |
11 |
0 |
9 |
5 |
3 |
7 |
6 |
2 |
10 |
4 |
| [5]Ri bit positions |
5 |
8 |
7 |
0 |
3 |
2 |
11 |
4 |
13 |
6 |
1 |
10 |
12 |
9 |
| [6]Ri bit positions |
8 |
9 |
5 |
13 |
0 |
10 |
7 |
1 |
12 |
3 |
2 |
4 |
11 |
6 |
| [7]Ri bit positions |
11 |
10 |
0 |
7 |
2 |
9 |
8 |
1 |
5 |
3 |
6 |
4 |
12 |
13 |
| [8]Ri bit positions |
11 |
4 |
0 |
13 |
10 |
12 |
5 |
7 |
2 |
8 |
3 |
1 |
6 |
9 |
| [9]Ri bit positions |
4 |
0 |
5 |
1 |
12 |
2 |
10 |
3 |
13 |
9 |
6 |
11 |
8 |
7 |
| [10]Ri bit positions |
4 |
7 |
0 |
8 |
10 |
1 |
6 |
3 |
2 |
9 |
11 |
12 |
13 |
5 |
| [11]Ri bit positions |
4 |
6 |
0 |
13 |
12 |
1 |
11 |
2 |
8 |
3 |
10 |
7 |
9 |
5 |
| [12]Ri bit positions |
0 |
5 |
1 |
9 |
2 |
12 |
3 |
6 |
8 |
7 |
4 |
10 |
11 |
13 |
| [13]Ri bit positions |
12 |
4 |
2 |
11 |
10 |
1 |
13 |
6 |
0 |
9 |
3 |
8 |
5 |
7 |
| [14]Ri bit positions |
10 |
6 |
0 |
13 |
12 |
11 |
8 |
5 |
2 |
4 |
3 |
1 |
9 |
7 |
| [15]Ri bit positions |
7 |
6 |
0 |
1 |
10 |
3 |
9 |
4 |
2 |
5 |
8 |
11 |
12 |
13 |
[0054] Bit permutation for the 32K mode
Receiver
[0055] Figure 7 provides an example illustration of a receiver which may be used with the
present technique. As shown in Figure 7, a COFDM signal is received by an antenna
300 and detected by a tuner 302 and converted into a digital form by an analogue-to-digital
converter 304. A guard interval removal processor 306 removes the guard interval from
a received COFDM symbol, before the data is recovered from the COFDM symbol using
a Fast Fourier Transform (FFT) processor 308 in combination with a channel estimator
and correction 310 in co-operation with a embedded-signalling decoding unit 311, in
accordance with known techniques. The demodulated data is recovered from a mapper
312 and fed to a symbol de-interleaver 314, which operates to effect the reverse mapping
of the received data symbol to re-generate an output data stream with the data de-interleaved.
[0056] The symbol de-interleaver 314 is formed from a data processing apparatus as shown
in Figure 7 with an interleaver memory 540 and an address generator 542. The interleaver
memory is as shown in Figure 4 and operates as already explained above to effect de-interleaving
by utilising sets of addresses generated by the address generator 542. The address
generator 542 is formed as shown in Figure 8 and is arranged to generate corresponding
addresses to map the data symbols recovered from each COFDM sub-carrier signals into
an output data stream.
[0057] The remaining parts of the COFDM receiver shown in Figure 7 are provided to effect
error correction decoding 318 to correct errors and recover an estimate of the source
data.
[0058] One advantage provided by the present technique for both the receiver and the transmitter
is that a symbol interleaver and a symbol de-interleaver operating in the receivers
and transmitters can be switched between the 1k, 2k, 4k, 8k, 16k and the 32k mode
by changing the generator polynomials and the permutation order. Hence the address
generator 542 shown in Figure 8 includes an input 544, providing an indication of
the mode as well as an input 546 indicating whether there are odd/even COFDM symbols.
A flexible implementation is thereby provided because a symbol interleaver and de-interleaver
can be formed as shown in Figures 3 and 8, with an address generator as illustrated
in either of Figures 5. The address generator can therefore be adapted to the different
modes by changing to the generator polynomials and the permutation orders indicated
for each of the modes. For example, this can be effected using a software change.
Alternatively, in other embodiments, an embedded signal indicating the mode of the
DVB-T2 transmission can be detected in the receiver in the embedded-signalling processing
unit 311 and used to configure automatically the symbol de-interleaver in accordance
with the detected mode.
[0059] Various modifications may be made to the embodiments described above without departing
from the scope of the present invention. In particular, the example representation
of the generator polynomial and the permutation order which have been used to represent
aspects of the invention are not intended to be limiting and extend to equivalent
forms of the generator polynomial and the permutation order.
[0060] As will be appreciated the transmitter and receiver shown in Figures 1 and 7 respectively
are provided as illustrations only and are not intended to be limiting. For example,
it will be appreciated that the position of the symbol interleaver and the de-interleaver
with respect, for example to the bit interleaver and the mapper can be changed. As
will be appreciated the effect of the interleaver and de-interleaver is unchanged
by its relative position, although the interleaver may be interleaving I/Q symbols
instead of v-bit vectors. A corresponding change may be made in the receiver. Accordingly
the interleaver and de-interleaver may be operating on different data types, and may
be positioned differently to the position described in the example embodiments.
[0061] According to an example implementation of a receiver there is provided a data processing
apparatus which is operable to map symbols received from a predetermined number of
sub-carrier signals of an Orthogonal Frequency Division Multiplexed (OFDM) symbol
into an output symbol stream. The data processing apparatus comprises a de-interleaver
operable to read-into a memory the predetermined number of data symbols from the OFDM
sub-carrier signals, and to read-out of the memory the data symbols into the output
symbol stream to effect the mapping. The read-out is in a different order than the
read-in, the order being determined from a set of addresses, with the effect that
the data symbols are de-interleaved from the OFDM sub-carrier signals. An address
generator is operable to generate the set of addresses, an address being generated
for each of the received data symbols to indicate the OFDM sub-carrier signal from
which the received data symbol is to be mapped into the output symbol stream. The
address generator comprises a linear feedback shift register, which includes a predetermined
number of register stages and is operable to generate a pseudo-random bit sequence
in accordance with a generator polynomial. A permutation circuit is operable to receive
the content of the shift register stages and to permute the bits present in the register
stages in accordance with a permutation order to form an address of one of the OFDM
sub-carriers, and a control unit operable in combination with an address check circuit
to re-generate an address when a generated address exceeds a predetermined maximum
valid address. The data processing apparatus is characterised in that the predetermined
maximum valid address is approximately thirty two thousand, the linear feedback shift
register has fourteen register stages with a generator polynomial for the linear feedback
shift register of

and the permutation order forms, with an additional bit, a fifteen bit address
Ri[
n] for the i-th data symbol from the bit present in the n-th register stage

in accordance with the table:
| R'i bit positions |
13 |
12 |
11 |
10 |
9 |
8 |
7 |
6 |
5 |
4 |
3 |
2 |
1 |
0 |
| Ri bit positions |
6 |
5 |
0 |
10 |
8 |
1 |
11 |
12 |
2 |
9 |
4 |
3 |
13 |
7 |
[0062] As explained above the permutation codes and generator polynomial of the interleaver,
which has been described with reference to an implementation of a particular mode,
can equally be applied to other modes, by changing the predetermined maximum allowed
address in accordance with the number of sub-carriers for that mode.
[0063] As mentioned above, embodiments of the present invention find application with DVB
standards such as DVB-T, DVB-T2 and DVB-H, which are incorporated herein by reference.
For example embodiments of the present invention may be used in a transmitter or receiver
operating in accordance with the DVB-H standard, in handheld mobile terminals. The
mobile terminals may be integrated with mobile telephones (whether second, third or
higher generation) or Personal Digital Assistants or Tablet PCs for example. Such
mobile terminals may be capable of receiving DVB-H or DVB-T compatible signals inside
buildings or on the move in for example cars or trains, even at high speeds. The mobile
terminals may be, for example, powered by batteries, mains electricity or low voltage
DC supply or powered from a car battery. Services that may be provided by DVB-H may
include voice, messaging, internet browsing, radio, still and/or moving video images,
television services, interactive services, video or near-video on demand and option.
The services might operate in combination with one another.
[0064] In other examples embodiments of the present invention finds application with the
DVB-T2 standard as specified in accordance with ETSI standard EN 302 755. In other
examples embodiments of the present invention find application with the cable transmission
standard known as DVB-C2. However, it will be appreciated that the present invention
is not limited to application with DVB and may be extended to other standards for
transmission or reception, both fixed and mobile.
1. A data processing apparatus operable to map input data symbols to be communicated
onto a predetermined number of sub-carrier signals of an Orthogonal Frequency Division
Multiplexed (OFDM) symbol, the data processing apparatus comprising
an interleaver operable to read-into a memory the predetermined number of data symbols
for mapping onto the OFDM sub-carrier signals, and to read-out of the memory the data
symbols for the OFDM sub-carriers to effect the mapping, the read-out being in a different
order than the read-in, the order being determined from a set of addresses, with the
effect that the data symbols are interleaved on the sub-carrier signals,
an address generator operable to generate the set of addresses, an address being generated
for each of the input symbols to indicate one of the sub-carrier signals onto which
the data symbol is to be mapped, the address generator comprising
a linear feedback shift register including a predetermined number of register stages
and being operable to generate a pseudo-random bit sequence in accordance with a generator
polynomial,
a permutation circuit operable to receive the content of the shift register stages
and to permute the bits present in the register stages in accordance with a permutation
order to form an address of one of the OFDM sub-carriers, and
a control unit operable in combination with an address check circuit to re-generate
an address when a generated address exceeds a predetermined maximum valid address,
characterised in that
the predetermined maximum valid address is approximately thirty two thousand,
the linear feedback shift register has fourteen register stages with a generator polynomial
for the linear feedback shift register of

and the permutation order forms, with an additional bit, a fifteen bit address Ri[n] for the i-th data symbol from the bit present in the n-th register stage

in accordance with the table:
| R'i bit positions |
13 |
12 |
11 |
10 |
9 |
8 |
7 |
6 |
5 |
4 |
3 |
2 |
1 |
0 |
| Ri bit positions |
6 |
5 |
0 |
10 |
8 |
1 |
11 |
12 |
2 |
9 |
4 |
3 |
13 |
7 |
2. A data processing apparatus as claimed in Claim 1, wherein the predetermined maximum
valid address is a value substantially between twenty four thousand and thirty two
thousand seven hundred and sixty eight.
3. A data processing apparatus as claimed in Claim 1, wherein the OFDM symbol includes
pilot sub-carriers, which are arranged to carry known symbols, and the predetermined
maximum valid address depends on a number of the pilot sub-carrier symbols present
in the OFDM symbol.
4. A data processing apparatus as claimed in Claim 1, 2 or 3, wherein the interleaver
memory is operable to effect the mapping of the input data symbols onto the sub-carrier
signals for even OFDM symbols by reading in the data symbols according to the set
of addresses generated by the address generator and reading out in a sequential order,
and for odd OFDM symbols by reading in the symbols into the memory in a sequential
order and reading out the data symbols from the memory in accordance with the set
of addresses generated by the address generator.
5. A transmitter for transmitting data using Orthogonal Frequency Division Multiplexing
(OFDM), the transmitter including a data processing apparatus according to any preceding
Claim.
6. A transmitter as claimed in Claim 5, wherein the transmitter is operable to transmit
data in accordance with a Digital Video Broadcasting standard such as the Digital
Video Broadcasting-Terrestrial, Digital Video Broadcasting-Handheld standard or the
Digital Video Broadcasting-Terrestrial2 standard.
7. A method of mapping input data symbols to be communicated onto a predetermined number
of sub-carrier signals of an Orthogonal Frequency Division Multiplexed (OFDM) symbol,
the method comprising
reading-into a memory the predetermined number of data symbols for mapping onto the
OFDM sub-carrier signals,
reading-out of the memory the data symbols for the OFDM sub-carriers to effect the
mapping, the read-out being in a different order than the read-in, the order being
determined from a set of addresses, with the effect that the data symbols are interleaved
on the sub-carrier signals,
generating the set of addresses, an address being generated for each of the input
symbols to indicate one of the sub-carrier signals onto which the data symbol is to
be mapped, the generating the set of addresses comprising
using a linear feedback shift register including a predetermined number of register
stages to generate a pseudo-random bit sequence in accordance with a generator polynomial,
using a permutation circuit operable to receive the content of the shift register
stages to permute the bits present in the register stages in accordance with a permutation
order to form an address, and
re-generating an address when a generated address exceeds a predetermined maximum
valid address, characterised in that
the predetermined maximum valid address is approximately thirty two thousand,
the linear feedback shift register has fourteen register stages with a generator polynomial
for the linear feedback shift register of

and the permutation order forms, with an additional bit, a fifteen bit address Ri[n] for the i-th data symbol from the bit present in the n-th register stage

in accordance with the table:
| R'i bit positions |
13 |
12 |
11 |
10 |
9 |
8 |
7 |
6 |
5 |
4 |
3 |
2 |
1 |
0 |
| Ri bit positions |
6 |
5 |
0 |
10 |
8 |
1 |
11 |
12 |
2 |
9 |
4 |
3 |
13 |
7 |
8. A method as claimed in Claim 7, wherein the predetermined maximum valid address is
a value substantially between twenty four thousand and thirty two thousand seven hundred
and sixty eight.
9. A method as claimed in Claim 7, wherein the OFDM symbol includes pilot sub-carriers,
which are arranged to carry known symbols, and the predetermined maximum valid address
depends on a number of the pilot sub-carrier symbols present in the OFDM symbol.
10. A method of transmitting data using Orthogonal Frequency Division Multiplexing (OFDM),
the method including
receiving data symbols to be communicated onto a predetermined number of sub-carrier
signals of an Orthogonal Frequency Division Multiplexed (OFDM) symbol,
reading-into a memory the predetermined number of data symbols for mapping onto the
OFDM sub-carrier signals,
reading-out of the memory the data symbols for the OFDM sub-carriers to effect the
mapping, the read-out being in a different order than the read-in, the order being
determined from a set of addresses, with the effect that the data symbols are interleaved
on the sub-carrier signals,
generating the set of addresses, an address being generated for each of the input
symbols to indicate one of the sub-carrier signals onto which the data symbol is to
be mapped, the generating the set of addresses comprising
using a linear feedback shift register including a predetermined number of register
stages to generate a pseudo-random bit sequence in accordance with a generator polynomial,
using a permutation circuit operable to receive the content of the shift register
stages to permute the bits present in the register stages in accordance with a permutation
order to form an address, and
re-generating an address when a generated address exceeds a predetermined maximum
valid address, and
transmitting the data symbols read-out of the memory using the OFDM sub-carriers,
characterised in that
the predetermined maximum valid address is approximately thirty two thousand,
the linear feedback shift register has fourteen register stages with a generator polynomial
for the linear feedback shift register of

and the permutation order forms, with an additional bit, a fifteen bit address Ri[n] for the i-th data symbol from the bit present in the n-th register stage

in accordance with the table:
| R'i bit positions |
13 |
12 |
11 |
10 |
9 |
8 |
7 |
6 |
5 |
4 |
3 |
2 |
1 |
0 |
| Ri bit positions |
6 |
5 |
0 |
10 |
8 |
1 |
11 |
12 |
2 |
9 |
4 |
3 |
13 |
7 |
11. A method of transmitting as claimed in Claim 10, wherein the transmitting the data
symbols read-out of the memory using the OFDM sub-carriers is in accordance with a
Digital Video Broadcasting standard such as the Digital Video Broadcasting-Terrestrial,
Digital Video Broadcasting-Handheld standard or the Digital Video Broadcasting-Terrestrial2
standard.
12. An address generator for use with transmission of data symbols interleaved onto sub-carriers
of an Orthogonal Frequency Division Multiplexed symbol, the address generator being
operable to generate a set of addresses, each address being generated for each of
the data symbols to indicate one of the sub-carrier signals onto which the data symbol
is to be mapped, the address generator comprising
a linear feedback shift register including a predetermined number of register stages
and being operable to generate a pseudo-random bit sequence in accordance with a generator
polynomial,
a permutation circuit operable to receive the content of the shift register stages
and to permute the bits present in the register stages in accordance with a permutation
order to form an address, and
a control unit operable in combination with an address check circuit to re-generate
an address when a generated address exceeds a predetermined maximum valid address,
characterised in that
the predetermined maximum valid address is approximately thirty two thousand,
the linear feedback shift register has fourteen register stages with a generator polynomial
for the linear feedback shift register of

and the permutation order forms, with an additional bit, a fifteen bit address Ri[n] for the i-th data symbol from the bit present in the n-th register stage

in accordance with the table:
| R'i bit positions |
13 |
12 |
11 |
10 |
9 |
8 |
7 |
6 |
5 |
4 |
3 |
2 |
1 |
0 |
| Ri bit positions |
6 |
5 |
0 |
10 |
8 |
1 |
11 |
12 |
2 |
9 |
4 |
3 |
13 |
7 |